"Voice has the potential to be the most naturalistic technology we have ever invented."
James Vlahos

Article: The emerging type of human-machine interaction.

Our language is what really sets us apart as a species. Words define and connect us. Teaching machines human language has never been more promising to bridge the gap between humans and machines and replace the predominant and unnatural devices used to control computers.

With SEMMI - the multilingual, scalable and learning AI concierge solution - Deutsche Bahn sets its stage for the age of voice assistants. The Conversational AI solution SEMMI receives user requests in natural language, analyses, understands and fulfills them by returning appropriate answers and executing tasks.

Conversational user interfaces for a convenient user interaction.

Be it a digital avatar, a humanoid service robot or an ancient telephone: The user interacts with SEMMI via voice-capable user interfaces in his natural language. SEMMI receives the user's request as an audio file, ready to apply the latest achievements of voice technology to create a personalized experience for the user. Choose SEMMI Avatar to combine the most convenient input methods from both worlds: voice and touch.


Utilising artificial intelligence to recognise and understand natural language.

Every spoken request to SEMMI is initially converted into text ("Speech To Text" or STT). Transcribing spoken words into text may be a no-brainer for humans, yet it is fairly difficult for computers, being originally built to process numbers. Once this challenge is solved, SEMMI uses further AI capabilities ("Natural Language Understanding" or NLU) to identify the intent of the user's request and thus being able to process it.


Accessing, retrieving and preparing information.

Whether a user asks about train directions, the way to the closest coffee vendor, your service portfolio or the weather: SEMMI connects to your internal knowledge repositories as well as third-party services via application programming interfaces (APIs), retrieves information and tailors a suitable response using natural language, facial expressions and gestures.


Conversations in natural language.

By converting "text to speech" (speech synthesis), SEMMI returns the answer to the user via the conversational user interface. Depending on the capabilities of the device, SEMMI answers the user request via speech, visualisation as well as facial expressions and gestures to enhance the conversation emotionally with a smile or a wink.


Article: Learn how SEMMI can help you to engage your customers and relieve your service staff.

Demo vereinbaren

Solution-oriented service

From implementing single core components to a fully-managed setup including hardware lease, infrastructure operations, repository integrations and utilization assessments, SEMMI completely engages with the scope of your voice assistant project. Thus, operators of SEMMI benefit from a solutions tailored to the specific requirements of each individual use scenario.

Focus on user-centric outcomes

We focus on orchestrating and leveraging well established AI technology from market leading vendors instead of developing own proprietary natural language processing engines. This enables us to focus on the real competitive advantage: the underlying knowledge base and training pipelines to maximise the quality of the user dialogue.

No vendor lock-in

With the flexibility of an architecture that allows to change specific technological components for each individual use case, SEMMI avoids dependencies on any technology vendor at this early stage of conversational AI.

A highly integrative voice first setup

Your first experience with virtual assistants will likely not be the last one. Thus, arranging your business data to integrate with SEMMI AI becomes a valuable asset for further conversational AI projects. Platforms like SEMMI AI support you to build the basis for multiple virtual assistants serving different audiences and use cases.

100 % GDPR compliant

To operate several consumer-facing conversational AI setups at Deutsche Bahn, data privacy and data protection is one of the most crucial requirements to meet. Be it on premise or a cloud-service architecture, together with your data privacy officers we ensure that GDPR regulations are being met.

SEMMI Avatar

The smartest multimodal human-machine interaction your customers ever experienced.

Learn more



The socio-empathic, humanoid service robot with a guaranteed wow experience for your customers.

Learn more


SEMMI Hotline

The smart interactive voice response system (IVR) to frictionless scale your customer contact centers.

Learn more


Other conversational user interfaces

SEMMI AI is built with an accessible API to operate any voice-controlled device of your choice - be it robots, digital avatars, mobile devices, smart speakers or others.